Smart Droplets Knowledge Base

Welcome to the Smart Droplets Knowledge Base!

Here you will find the Smart Droplet Academy's resources available on-demand.

The Smart Droplets Academy is an initiative designed to provide education and training on cutting-edge technologies with applications in the agricultural sector. The Academy is part of the Smart Droplets project, which aims to develop an intelligent crop spraying system.

The Academy offers a variety of learning opportunities that cover topics such as:

The Knowledge Base is divided into dedicated sections of seminars organised by each Smart Droplets partner. These sections feature a wide array of educational resources, including webinars, presentations, case studies, code notebooks, and other materials. Each partner brings unique expertise, allowing participants to explore innovations and technologies tailored to the agricultural sector.

The webinars, workshops, and seminars are designed for a broad audience, including students, farmers, researchers, and IT professionals. The instructors are experts from leading universities and research institutions.

The Academy’s goal is to empower stakeholders in the agri-food sector with the knowledge and skills they need to adopt new technologies, improve agricultural practices, and contribute to building a more sustainable and efficient future for farming.

Below, you’ll find resources for the Academy courses, including short descriptions of each course, video lectures, presentations, and code notebooks with the code examples provided in the lectures.

This series of lectures explores the application of AI in agricultural modeling, with a specific focus on reinforcement learning and digital twins. We delve into how AI can be integrated with traditional process-based models to enhance decision-making and predictive capabilities.

Key topics include:

Digital Twins in Agriculture: Creating virtual representations of real-world agricultural systems to simulate and optimize various scenarios.
Crop Growth Modeling with AI: Utilizing AI to improve the accuracy and precision of crop growth models, leading to better yield predictions and resource management.
Reinforcement Learning for Agricultural Decision Making: Implementing reinforcement learning algorithms to learn optimal decision-making policies for agricultural tasks such as fertilization.

Introduction to Smart Droplets and Reinforcement Learning

Seminar on Reinforcement Learning and CropGym

Lecture on Reinforcement Learning

Reinforcement Learning for Nitrogen Fertilization Recommendations

Workshop on Digital Twins

Lecture on crop growth modeling with AI

Introduction to Machine Learning

Image Classification: From Fully Connected Networks to CNNs. A Weed Detection Approach.

Image Classification: From CNNs to Transformers. A use case for Weed Identification.

Transfer Learning for small datasets in Weed Identification

Object Detection (Faster R-CNN and YOLO)

Annotations Strategies and Techniques/Intro to Object Detection

In this comprehensive webinar series, presented by robotics experts from Eurecat, we dive deep into the engineering and software powering the future of mobile robotics in precision agriculture. Moving from foundational concepts to real-world deployment, this series is designed to bridge the “Reality Gap” between ideal simulations and the challenging, dynamic environments of real farms.

Throughout this series, we will:

Master the Fundamentals: Explore the Robot Operating System (ROS) architecture, configure nodes, and learn to leverage essential libraries like OpenCV and MoveIt to accelerate robot development.
Navigate the “Green Desert”: Discover why standard GPS is insufficient for farming and learn how advanced Simultaneous Localization and Mapping (SLAM) algorithms solve complex visual and seasonal challenges.
Deconstruct Robotic Autonomy: Dive into the decision-making psychology of autonomous robots, exploring the cognitive architectures and algorithms needed to navigate non-geometric terrain and organic obstacles.
Tackle Real-World Engineering: Get a behind-the-scenes look at retrofitting heavy tractors, overcoming practical hurdles like mud, sensor vibration, and signal loss through 3D LiDAR perception and multi-stage sensor fusion.

Introduction to mobile robotics with ROS

Simultaneous Localization and Mapping (SLAM)

Planning and Navigation

Lessons Learned

In this webinar series, we explore the cutting-edge application of Artificial Intelligence in agricultural modeling. Join us as we dive deep into how AI can be integrated with traditional process-based models to enhance decision-making, boost predictive capabilities, and shape the future of smart farming.

Throughout this series, we will cover key topics including:

Digital Twins in Agriculture: Learn how to create virtual representations of real-world farming systems to safely simulate and optimize various agricultural scenarios.
Crop Growth Modeling with AI: Discover how utilizing Artificial Intelligence improves the accuracy and precision of crop growth models, leading to better yield predictions and smarter resource management.
Reinforcement Learning: Understand how to implement advanced reinforcement learning algorithms to develop optimal decision-making policies for critical agricultural tasks, such as precise fertilization.